The present study aimed to explore the knowledge and thoughts of hospital physicians representing different specialties of the DNAR process. Our results demonstrated specialty-related variation in understanding of the meaning of a DNAR order as well as perceived challenges in making the decision and documentation of the order. These issues may have negative effects on the treatment of patients with severe or irreversible conditions.
Healthcare professionals consider making a DNAR order as challenging [1, 2]. Placing a DNAR order is associated with complex medical, psychosocial, ethical and emotional aspects [10]. For clinicians, making a DNAR order includes assessment of several clinical issues, such as prognosis, likely outcomes of CPR, and patient comorbidities [11]. In a previous Finnish study, physicians considered their competence in making DNAR decisions to be sufficient [1]. In the present study, almost half of physicians in conservative and operative specialties, psychiatry and pediatrics perceived DNAR order making as challenging or difficult, whereas only 25–31% of anesthesiologists and oncologists found such problems. The experience of oncologists with terminal illnesses and anesthesiologists’ work in ICUs with acute and severe disorders probably helps in making difficult decisions.
Depending on the specialty, 23–91% of physicians considered that the opinions of patients and/or family should be involved in the DNAR order process. According to Finnish legislation, all care-related decisions, such as DNAR order-making, should be discussed with the patient or, if the patient is incompetent, with relatives. National legislation, however, gives no specific guidance for making a DNAR order [12]. The National Supervisory Authority for Welfare and Health states that the DNAR order is a medical decision made by the physician responsible for the care of the patient [13]. A competent patient can refuse any treatment, including life-sustaining treatment, but he/she cannot demand a treatment, such as CPR, that a physician considers not to be in the person’s best interests [11]. However, in a Swedish study, 16% of physicians stated that a patient’s wish to receive CPR should always be respected [14]. In the same study, 8% of physicians considered that relatives could make the decision [14].
A previous Finnish study found that 70% of physicians considered that a DNAR order forbade CPR only, but 30% of respondents believed that it implied palliative care [1]. In our study, there was also considerable variation in opinion on what a DNAR entails. Overall, almost two thirds of physicians in oncology but less than 50% of respondents in the largest specialties (anesthesiology/ICU, conservative and operative specialties) correctly recognized that the DNAR order excludes only CPR. Better knowledge among oncologists is probably due to their work with patients with severe or terminal illness and appropriate training. Of respondents working in adult specialties, 42–63% considered that care in the ICU would also be excluded. Misinterpretation of the meaning of a DNAR order and challenges implementing it into practice is common among physicians [2–4, 15]. In a study in Ireland, 21% of hospital doctors considered that patients with a DNAR order could not receive chemotherapy, radiotherapy or dialysis, and they could not be referred for surgery. About 6% thought that these patients could not be administered antibiotics, intravenous fluids or nasogastric feeding [3]. In an UK study, half of physicians thought that a DNAR order alters how in practice fluids and pain relief are given to those patients [4]. The significance of the misinterpretation of a DNAR order is that it may lead to suboptimal treatment of patients [4, 16].
Most responders in our study considered that DNAR orders are made at an adequate time in relation to the condition of the patient. However, according to 60–80% of physicians, decisions are being made after office hours. Thus, in most cases a DNAR order is likely to be made by another doctor than the one who is overall responsible for the treatment of the patient. Timely decision of a DNAR order and its documentation may be hampered for various reasons, such as unexpected deterioration of a patient, delay in discussions with the patient and relatives, or consultations with other colleagues [5].
In order to act correctly according to previously made decisions it is essential that valid DNAR orders are clearly recorded and readily available. A systematic review found that DNAR orders are often inconsistently recorded [15]. CPR is infrequently included in hospital orders or medical records of patients who undergo CPR during their hospital stay [6, 7, 17]. Furthermore, the frequency of documented DNAR orders varies depending on the type of patient admission and the treating department: documentation is more common in acute than planned admissions, and DNAR orders are most commonly made at the department of internal medicine and pulmonology [6]. In the era of electronic patient records (EPR), DNAR orders should be easily available. This requires, however, that the order is documented appropriately and in a standardized manner. According to the Finnish National Institute for Health, a DNAR order should be recorded in the risk information part of the EPR by a doctor [18]. The large majority of our respondents correctly recognized this guidance. However, 40–62% of physicians of different specialties experienced challenges with the documentation of DNAR orders.
Overall, physicians’ confidence regarding DNR discussions is low compared to their confidence regarding other medical discussions [19]. Several studies have pointed out the need for improvement of education in DNAR decision making [2, 3, 8]. Many physicians report that they have learned communication skills relating to DNAR orders and end-of-life discussions from their clinical experience [20]. Lack of adequate training in the DNAR order process and its implementation is likely the main reason for the considerable variation among specialties in perceived adequacy of training in DNAR order making. Lowest rates were observed in pediatrics, pediatric neurology and psychiatry. This may reflect the views of both educators and trainees in these fields that CPR and/or DNAR are not common issues in their field. Efforts to improve DNR discussions should explore the need to tailor educational interventions to fit these characteristics.
A strength of our study is the participation of a rather large number of physicians representing a wide range of medical specialties in several secondary/tertiary care hospitals. Thus, because there may be differences in DNAR policies and practices among hospitals, we believe that our study provides a representative illustration of how a DNAR order is made in specialized care in Finland. Our study has some limitations. We do not know the actual response rate as the exact number of physicians working in the index hospitals at the time of survey is not available. This is because some physicians may have been on vacation, sick leave or sabbatical. In addition, because responses were returned anonymously, we were not able to contact subjects who did not answer the questionnaire. Furthermore, the questionnaire was not pretested.
Conclusions: In conclusion, our findings show specialty-related variation among hospital physicians in the interpretation of the scope of a DNAR order and who should be included in the decision-making process. Furthermore, even in the era of EPRs, problems with the documentation of DNAR orders still exist. All of these issues may compromise the care of patients. Our study also indicates that many hospital physicians are in need of more education and training in end-of-life decision making. Finally, in addition to national and international guidelines, there is a need for clear legislation.